Predicting common maternal postpartum complications: leveraging health administrative data and machine learning

Abstract

Objective We aimed to predict the risk of common maternal postpartum complications requiring an inpatient episode of care. Design and setting Maternal data from the beginning of gestation up to and including the delivery, and neonatal data recorded at delivery, were used to predict postpartum complications. Sample Administrative health data of all inpatient live births (n = 422 509) in the Australian state of Queensland between January 2009 and October 2015. Method Gradient boosted trees were used with five‐fold cross‐validation to compare model performance. The best performing models for each outcome were then assessed in the independent validation data using the area under the receiver operating curve (AUC‐ROC). Main outcome measure Postpartum complications occurring in the first 12 weeks after delivery requiring hospital admission. Results Postpartum hypertensive disorders obtained good discrimination in the independent validation data (AUC = 0.879, 95% CI 0.846–0.912), as did obstetric surgical wound infection (AUC = 0.856, 95% CI 0.838–0.873), whereas postpartum sepsis and haemorrhage obtained poor discrimination. Conclusions Our study suggests that routinely collected health data have the potential to play an important role in helping determine women's risk of common postpartum complications leading to hospital admission. This information can be presented to clinical staff after delivery to help guide immediate postpartum care, delayed discharge, and post‐discharge patient follow up. For such a system to be effective and valued, it must produce accurate predictions, and our findings suggest areas where routine data collection could be strengthened to this end. Tweetable abstract Improved prediction of maternal postnatal hypertensive disorders and wound infection via machine learning.

Artificial Intelligence in Pregnancy: A Scoping Review

“…• Maternal anemia was addressed using ML models by [173] and [174]. • Postpartum complications were reported by [175] (health administrative data and ML techniques were applied to predict the risk of common maternal postpartum complications) and [52] (DM models were used to create to predict the need for neonatal resuscitation)…”

Prediction of Maternal Hemorrhage: Using Machine Learning to Identify Patients at Risk

“…Postpartum hemorrhage is a known cause of significant maternal morbidity and mortality in the United States and remains difficult to predict. Few existing studies have utilized machine learning methods to identify patients at risk for postpartum hemorrhage with minimal success 5,6,7,8 . A recently published model used a large cohort from the U.S Consortium for Safe Labor and achieved excellent discrimination, although its utility in the clinical setting is limited 11 .…”

“…Previous models for prediction of postpartum hemorrhage have been developed 5,6,7 , but validation of these among different populations and at different time points within the labor process has been limited. A machine learning study using administrative data provided poor discrimination for predicting need for hospital readmission due to postpartum in the first 12 weeks postpartum 8 . Prediction of postpartum hemorrhage remains a challenge for the obstetric provider and further work is necessary using modern modeling methods.…”

Artificial intelligence in obstetrics

“…From January 2009 to October 2015, machine learning analysis of predictions of postpartum complications using data from all live births (n=422,509) in Queensland, Australia showed hypertension disorder (AUC, 0.879; 95% CI, 0.846-0.912) and postpartum wound infections (AUC, 0.856; 95% CI, 0.838-07). The authors argued that usage of such methods could play an important role in determining the risk of postpartum complications in advance [56]. Shoulder dystocia must be the least accessible clinical situation for obstetricians.…”